Process Monitoring and Estimation of Material Properties of Additively Manufactured Components Using Model-Based Inversion of Process Compensated Resonance Testing Data

Abstract

Abstract Certification of additively manufactured (AM) parts and qualification of AM processes, suppliers, and machines for aerospace applications involve significant mechanical testing, cost, and time. The AM community requires capabilities for fast, affordable, and effective certification and qualification. These capabilities include rapid component and model validation, characterization of process and geometry impact on material properties, build and process monitoring, and effective nondestructive evaluation (NDE). Process compensated resonance testing (PCRT) uses the resonance frequencies of a component to rapidly detect defective parts, monitor build and processes and post-processing, characterize material properties, and can be used as a basis for model validation. This study examines the use of PCRT model-based inversion to nondestructively determine the material properties of as-built titanium AM components. Over 100 samples were printed with intentional variations in build parameters, with the intention to produce generic “acceptable” and “unacceptable” components. PCRT modeling tools were then used to train the inversion code algorithms. The model-based inversion estimated the elastic properties of the as-built parts as well as a residual fit error, identifying several components with outlying properties. Parts were then mechanically tested, and the outliers were confirmed. As PCRT model-inversion is a powerful NDE method, several examples are described showing how inversion results can augment traditional AM inspection techniques.